The EP-TS and EP as a discipline

A few weeks ago, I raised a question on the CEP forum: what exactly is CEP? Everyone seems to agree that CEP is part of event-driven computing, but my question went to exactly how CEP differs from other components of event-driven computing.

The question came from a bit of searching for CEP definitions while writing a Knol about CEP. On one hand, we have David Luckham’s definition of CEP in his glossary. But I wanted to look deeper and find more detailed descriptions of CEP. See the Knol for links that I found to definitions and the glossary. I found that every description of CEP (other than David’s) follows one of two formats:

1. “CEP a technology to derive knowledge from events” – but with no description of exactly how this knowledge is supposed to be derived.
2. “Here are some concepts from more established disciplines that can be used in CEP” – with plenty of descriptions of techniques to derive knowledge, but no unifying theory as to what CEP adds to these better established disciplines.

So I was looking for those things that are uniquely CEP, or where CEP ties together other disciplines in a well described way (rather than simply listing a bunch of stuff that could be a part of CEP). I wanted to find the “meat” of CEP, the stuff that someone would need to be able to dig in to the topic.

Well since that post, I’ve come to realize that there is some existing content that I think shows more progress in this area than I realized.

For those who are interested in Event Processing but are not following the Event Processing Technical Society or EP-TS, I’ll suggest looking over the files in their Document Management section. Under “Event Processing Use Cases”->”Use Case Workgroup” they’ve got some terminology and reference architecture documents that are a great start on finally pulling together EP into a form that users and researchers will understand.

Also, the posts from Opher’s blog over the past year contain many ways to tie existing fields together in a coherent way to form Event Processing.

So if a graduate student were looking for a nice little project, I think that a coherent integration of topics from Opher’s blog with those EP-TS documents would be both educational and a great contribution to the community.

In any case, in its short life the EP-TS has already made some progress in pulling together the discipline of EP. I hope they keep up the good work.